Game on or off? A model of resource pre-emption by plants reveals a novel mechanism of coexistence

Background/Question/Methods

Plant competition can be modeled using game theory. This allows plants to respond to both neighbours and resources, where traditionally it has been assumed plants only respond to resources. This shifting world view has provided novel insights into plant strategies for competition. Here, we ask how these strategies influence coexistence. We developed a consumer-resource model of competition for two substitutable soil resources. The model was constructed such that we could analyze it as an evolutionary game, or not (game-on or game-off). When the model is analyzed as game-off, plants select strategies based only upon the nutrient environment. When the model is analyzed as game-on, plants select strategies based on both nutrient environment and neighbour strategies. To examine model predictions, we first analyzed zero net-growth isoclines (ZNGIs) for game-on or game-off and discuss the consequences of these ZNGIs for species coexistence in either form of the model. Second, we explore the set of best responses plants use in either model in response to neighbour strategies, to ask how behavioural responses of plants influence coexistence. Third, we constructed random assemblages of hypothetical species and asked; how many species could coexist on two substitutable resources in either game-on or game-off.

Results/Conclusions

For game-off, the model reduces to a classic consumer-resource dynamic and in a two resource world only two species may coexist. Alternatively, for game-on, more than two species could coexist on two resources and any two species were more likely to coexist. This occurred because pre-emptive behaviour allowed plants in a game-on world to (i) subsidize their costs in a way that increased resource uptake and caused species mixtures to more completely use the resource base, and (ii) behavioural assessment and response to neighbour strategies acted like a focusing lens to line up the ZNGIs of multiple species through the depletion trajectory. Evidence is currently mixed, and we have a poor understanding of whether real plants are more commonly game-on or game-off, and thus whether this phenomena occurs. The best evidence in favour of a game-on world are recent discoveries of over-yielding of roots in mixtures compared to monocultures. The best evidence against a game-on world, are the plant species that do not seem to engage in pre-emptive games. This is potentially a novel mechanism of coexistence, and our comparison of game-on and game-off worlds generates alternative testable hypotheses for either possibility.